.

cJ

. ~.~.

FRill ENGINEERl'NG LABOR/HOP"

.~

~D2C

_'

LEHIGH UNIVE:RS1H
VETHLEHEMI.~ENNSYLVANIA

_....

'.

STRUCTURAL ANALYSIS
OF

I-IOT-METAL LADLES
by

KNUD-ENDRE KNUDSEN

An Abridgment of a Dissertation
Presented to the Graduate Faculty
of Lehigh University
in Candidacy for the Degree of
Doctor of Philosophy

-'

Reprinted from
IRON AND STEEL ENGINEER
Vol. XXVI

No. XII

December, 19-49

STRUCTURAL ANALYSIS
OF

J-IOT-METAL LADLES
by

KNUD-ENDRE KNUDSEN

An Abridgment of a Dissertation
Presented to the Graduate Faculty
of Lehigh University.
in Candidacy for the Degree of
Doctor of Philosophy

Reprinted from
IRON AND STEEL ENGINEER
Vol. XXVI

No. XII

December, 1949

STRESSES IN BOT METAL. LADLES

By K. E. KNUDSEN, WM. M. MUNSE and B. G. JOHNSTON

Fritz Engineering laboratory

lehigh University
Bethlehem, Pa.

REPRINTED FROM IRON AND STEEL ENGINEER, DECEMBER, 1949

STRESSES IN BOT METAL LADLES

By K. E. KNUDSEN, WM. H. MUNSE and B. G. JOHNSTON
,Fritz Engineering Laboratory
Lehigh University
Bethlehem, Pa.

. . . .this report describes the results of a

research investigation sponsored and
financed by the Association of Iron and
Steel Engineers for the steel industry ..

.. THE designation "Hot-Metal Ladle," when us~d in

this report, will mean a vessel, lined with refractory
material, used for conveying molten metal.
During the extensive growth of the steel industry
there has been a consistent trend to increase the
capacity of hot-metal ladles. Larger furnaces and thc
convenience of pouring the whole furnace charge into
one ladle have combined to make this increase necessary. At the same time the allowable load on existing
ladle cranes and supporting structures is limited, thereby rendering any decrease in ladle dead weight directly
applicable to increased capacity for molten metal. The
introduction of the welded type ladle in 1932 opened
new possibilities for decrease in dead weight (2). * Oval
shaped welded ladles came into use as an expedient to
increase capacity without interference with height
clearances and hook distances originally determined
for the round riveted ladles.
Thus, larger ladles are being built, new shapes
introduced, and the ratio of dead weight to ladle
capacity is forced down. This continual development
gives significance to the application of more rational
and accurate stress analysis procedures in order to
maintain the required safety and dependability of hotmetal ladles. Little design information is available in
the technical literature. The design procedures used by
the different ladle manufacturers apparently have given
completely satisfactory ladles. However, these procedures are in general based on assumptions which have
not necessarily been verified by tests.
A program of experimental and theoretical stress
analyses of hot-metal ladles was therefore adopted by
the Association of Iron and Steel Engineers as a part
of the Standardization Committee's postwar program.
Fritz Engineering Laboratory of Lehigh University
undertook the task of investigating the structural
behavior of such ladles, initial work starting on June 15,
Presented before AISE Annual Convention, Cleveland, Ohio, September 28, 1948

Reprinted from

IRON AND STEEL ENGINEER, DECEMBER, 1949

1946. The general program was determined at a committee meeting in August 1946. It was decided to test
3 models based on prototypes of 1.5Q-ton net capacity.
Both riveted and welded construction was considered,
and a model of an oval ladle was included. 'Additional
variables were: number and size of stiffener rings, size
of trunnions, angle of tilt, amount of load, and distance
between the points of support on the trunnion pins.
The important problems involving stresses due to
temperature differentials caused by the molten metal
are not considered in this investigation.
A progress report -was presented before the annual
convention of the AISE in Pittsburgh, September 1947.
The tests were completed in December 1947, and the
experimental program, procedures and results are described in detail in a separate test report (1). The
present report will therefore cover the conclusions only,
illustrated by typical experimental data.
The interest and advice given by Mr. Ingvald
Madsen, Research Engineer of AISE; Mr: F. E. Kling,
chairman of the ladle design committee, together with
other members of the committee, were essential factors
in the planning and execution of the program. The
valuable help of Mr. Paul Kaar, engineer of tests; and
Mr. Kenneth Harpel, foreman, is acknowledged, together with the valuable help of the many student
assistants in working up the strain gage data and preparing test result curves.
LADLE MODELS

The ladle types chosen for this investigation represent

three general types in use in the mills. They also reflect
different developments in ladle design. Ladle "A" is a
round riveted ladle, Ladle "B" is an oval welded ladle,
and Ladle "C" is a round welded type with a dished
bottom. All three specimens are 1/5 scale models of
Numb.r. refer to bibriography at end of report.

TABLE I
ACTUAL LADLE DIMENSIONS
Ladle shell

Size
in
tons

Ladle
shape

Ladle top
diameter, .
ft-in.

Type of
construction

Ladle bottom

- - - - - - - - - - - " - - - - - - - - - Method Date

Ladle
height,
ft-in.

Type of
trunnions

ThickShape ness, Brick,

in.
in.

Thickness, Brick,
in.
in.

Side
slope

of
pour

of
design

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - --1.25
12
25
45
50
75
75
100
120
120
130
150
150
150
150
150
190
200

Round
Round
Round
Round
Round
Round
Round
Elliptical
Elliptical
Round
Round
. Elliptical
Elliptical
Round
Elliptical
Elliptical
Elliptical
Elliptical

Riveted
Riveted
Welded
Riveted
Rivet.-Weld.
Riveted
Riveted
Rivet.-Weld.
Welded
Riveted
Riveted
Welded
Riveted
Riveted
Riveted
Rivet.-Weld.
Welded
Rivet.-Weld.

2-331

5-2
6-7
8-11%

8-431
9-6U

10-3
9-0 x 10-10%
9-7Ys x ll-llYs
11-3
11-9%
9-1031 x 12-7
11-6 x 13-6
12-7
11-4 x 12-6
11-7U x'13-231
11-5% x"14-5%
10-10 x 13-1031

2-4U

4
6
7U

U
%
%
%
%

Flat
Flat
Flat
Flat
Flat
Flat
Flat
Dish
Flat
Flat
Dish
Dish
Dish
Flat
Flat
Dish
Flat
Flat

......

6
Ys 8U
Ys 8-12
1% 7
1
5
1
8U
1% ......
1
6
lU 8
1
9
1
9
1% 7-931
1% 8
lU ~U-831

J16

Ys

Ys

1%
lU
1
lU

131

1
lU
1%
lU
1%
1%
1%

4
6
7U

....

10
12
15
9

831

12

....

12

1431

13

1331
1431
11
8

Top
Top
Top
Top
Bottom
Top
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom

1921
1921
1943
1926
1942
1916
1928
1944
1933
1916
1943
1940
1937
1929
1929
1943
1942
1944

from those of actual ladles approximately in the 5 to 1

ratio, although material availability was to some extent
a determining factor. The trunnion pins on all ladles

prototypes of 15Q.-ton net capacity. A survey of some

actual ladle designs is given in Table I for reference.
The dimensions and material thicknesses are reduced

Figure 1 -

%: 12
%: 12
%: 12
2*: 12
%: 12
1: 12
%: 12
%: 12
Ys: 12
1:12
1: 12
Ys: 12
Ys: 12
Ys: 12
H: 12
Ys: 12
1: 12
Ys: 12

Cast
Cast
Cast
Cast
Rivet.-Weld.
Cast
Cast
10-1031 Welded
11-4
Welded
11-0
Cast
11-3
Cast
Welded
12-431
11-4
Cast
Cast
13-2%
13-6
lCast
12-131 JRivet.-Weld.
12-1131 Welded
12-11
Welded

5-4
6-7
8-9
8-6
10-8
10-0

Ladle "A" is a one-fifth scale model of a lSD-ton riveted round ladle.

~L.

'j'# / /

It..
~ I
'j

"x'\ ~

/;/

-_.~--- ~e-~ - e-e-'- --:+

~"'" ~\::

.-

\ J..<..l.<-.. .- .1. -'~

-\\\ ~~

-f.-!-tr--~\I,
.L

11

/11)
'~~ "/1/\

~ ~

n~ w-

~~"----~:::---~
I

.....~-

.,,/'"

--=--=-~~

./0; rA.N1I

/011

,.,

;;l

.._nA_

/~x-

f--

"f-

"t

_4_

I~"'L

=--~
~,.lfIh

T. rJ JI,tT-c,. ri

.. rlr.........l-'Jl""*~
",.

- !-"

._~

-j:-

_..

I"f-

~:----"l~

"

Z'Pif'cJt &1

-j~~.

-----

"'---.
-

".,

,s1,1I__"",-.,

-1>

"oI~J ~o""M~II .."(_I,, loHId.';

fo ilt3itle .1' ""'")
/li"./c<l ('o,,,,eel/.It

Reprinted from

IRON AND STEEL ENGINEER, DECEMBER, 1949

were made long enough to permit study of the effect of

hook spreading, as required by the test program.
Some simplifications were made in the model design,
as compared with actual ladles. The slag spout and 'the
pouring device were eliminated. Special joints for facilitating the transport of the ladles were not considered.
The additional trunnion pins sometimes used when the
ladle is placed in stands were disregarded. It is believed
that the elimination of the slag spout at the lip of the
ladle is the only significant deviation of those here
mentioned. The test results indicate that the effect of
the lip ring on the structural b'ehavior of the ladle is
greater than emphasized in present design procedures.
, The serious discontinuity introduced in this ring by the
slag spout seems therefore to deserve special consideration. Tests with such spouts added on the models were
proposed at one step in the development of the test
program but were not adopted.
Ladle "A" is shown in Figure 1. Attention is called to
the 1/16 in. additional plate on the bottom, similar
plates on actual ladles being provided to protect against
heat radiation during the pouring operation. Ladle "A"
was tested with two different size trunnion pairs, also
with or without a stiffener band combined with the
small pair of trunnions. Figure 2 gives details of ladle
"B." The oval shape of the ladle is obtained by inserting
Figure 2 -

a straight middle section in the sides between the

semi-circular parts of the cross section. Ladle "B" has
a flat bottom with protection plate as in the case of
Ladle "A." On Ladle "C," Figure 3, the bottom is
dished and has no reinforcing plate. The stiffener rings
are of equal size and comparatively small due to the
heavier middle section of the shell between the rings.
The trunnions are built up of plates with no ribs of the
type used on Ladles "A" and "B," as may be seen from
Figures 4, 5, and 6.
To conform with general practice in ladle manufacturing, the models were made of structural carbon steel
conforming to ASTM Specification A-7 for heavier
material, and to ASTM Specification A 245-44T,
Grade C, for the thin plates. The models were stressrelieved for one half hour at 1150 F after fabrication.
Before testing, the ladles were lined with fireclay in a
way simulating the fire-brick lining used in actual
ladles. The thickness of the fireclay was %: in. on the
sides and 1 in. on the bottom, slightly more on Ladle
"A."
TEST PROGRAM AND RESULTS

In the various tests, strains and deflections were

measured at the points indicated in Figure 7 for Ladle

Ladle "8" is a one-fifth scale model of a 150-ton welded oval ladle.

ql----

GEN[RAL NOTES
Ladle to be sIre" relieved by

heating 10 uso-r, holdl"9 112

hOUri and slowly COOIlft9. All

trunnion Slru!s to b' 1/4e lhic:k

as shown. Welds to be in
aCCOrdance with the besl
practice and meeting AWS
'londOI'd. All dilnel'l$ions held

o. c;lose 01 possible. LI9ht

'<loge structural quollty flal hot
rolled corbon 11 1 conforming
10 ASTM speclfiC'Ofion A245~44T
GRADE C to bt used for thin

plot... Other port, to me,.

GRADE A 7 speclflcotlonl for
structural orod
1116- bottom .Ilff lng plo', 10M
bolted 10 bottom of lodl, wltl.
3/16"+bolt 3/64"tdrlll t

,.1.

boltom hoi

Reprinted from

IRON AND STEEL ENGINEltR, DECEMBER, 1949

"A." The locatIOns tor the other ladles were similar.

The strains were measured by means of bonded electric
wire resistance gages, mounted on identical locations
on the inside and outside ladle surfaces. The principal
stresses and their directions, as well as the horizontal
and vertical stress components, were computed from
the measured strains. Horizontal deflections of the sides
and vertical deflections of the bottoms were measured
by means of mechanical deflection dials reading to
1/1000 of one inch. Figure 8 gives an over-all picture of
the test set-up.
Actual ladles are loaded with molten metal weighing
at an average 420 lb per cu ft. Due to obvious inconveniences in handling molten metal, mercury was used
as loading agent in the laboratory tests.
A summary of the complete testing program is presented in Table II. The table lists the testing conditions
for all tests on each ladle, and indicates the measurements taken in each case. The variables included in
the test program may be found from a study of Table II,
and will be pointed out in connection with the discussion
of their effect on the structural behavior of the ladles.
The complete set of data is given in the unpublished
test report, as previously mentioned. The AISE may
be consulted if some of this data should be desired.
The general structural behavior of the ladle models

Figure 3 -

Ladle

lie"

deviates considerably from what is often assumed as a

basis for design. A description of this behavior will be
given before the effect of the test variables is discussed.
Figures 9 and 10 show representative deflection
results. They were obtained for ladle "A" under conditions explained in the figures. Figure 9 shows the
general tendency of the horizontal cross section to
become oval in shape due to the loading, the trunnion
regions moving inward, and the regions between trunnions moving outward. Zero deflection is found approximately 45 degrees from the trunnion lines. The side
deflections, as shown on vertical cross sections in
Figure 10, are zero at the bottom and increase nearly
linearly to the maxilpum at the lip. Some irregularity
is caused near the trunnion pin by the concentrated
hook reaction, especially for the larger hook distance.
The bottom deflection gives the picture of a partially
restrained circular plate under uniform load.
The stresses in the stiffener rings are primarily bending stresses, hence these stresses should be proportional
to the change in curvature which may be visualized in
Figure 9. This is confirmed by the stress measurements,
as shown in Figures 11 and 12, for the outside and inside
of the three stiffener rings on ladle "A." On each ring,
the maximum stresses occur at the trunnion line and
midway between the trunnions. The ring stresses are

is a one-fifth scale model of a 150-ton welded round ladle.

AII,.".I " ...... o~".,l" ... j'D-iI'~'

iao,. "' .

Jf

I.,.....

;.~~..N.

10

6"'FIlA~

#01' S
n .. ~ 1> .... olB Jot.A 'w/HI/. .... It> ~ ,...,.,..."-"J....'"

~I

""fl.

fII/U. C;oeJ.s f"

I ' .... ~ Rf,,!$ ~It_

""j. fIb.~ C'<i"I'r".-..A.. "" "H;" 01'

-'.)0. All n~h Jb be 4". r"lt""
1 .1..
".s.s I..".., .)' ~f'''V "".
1110-'. W"':"9 YZ "1N'.s o ..d JM"'7 .--1,,'9
~/~SS "hr>l';'~ I .t"." <JI~... _1"/~
It. '."
.11<1,.

1'0" s'..

.4/1

do",vu'''~

J,.~

..

""t ....

,,,,-11.1

, , _ , _ n"",
O~,.
Iw
,....r'
..6~,l"'ico.f,_. ,,,,. .s~J..,.1
,,..,,/. 1tI,'d.l ,""Iw", #'C .-d.~.
.,tM D..., p,... I,,;.. .,. . _.1,.."
AWS
.11"..

1'\0 H

g"",,~

9---" ".,

"".-.1,

c_

i-t'::J;;;~~~/~j~.l:9JI;;:~"itJ~;::~'::'Z
4'>''''

J,P<Ic.l;c"I,,,,,~

AHS-""r. t:i_c.

Reprinted from

IRON AND STEEL ENGINEER, DECEMBER, 1949

TABLE II
SUMMARY OF THE TEST PROGRAM
Variables
Entry
No.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
(1)
(2)
(3)
(4)

.
Model

Ladle

"A"

Ladle

"8"

Ladle

"e"

Loading
agent

Trunnion size,
in.

Mercury
Mercury
Mercury
Mercury
Mercury
Mercury
Mercury

8 x8~
8 x 87:4:
16 x 87:4:
16 x 87:4:
16 x 87:4:
16 x 8~
16 x 87:4:

Water
Mercury
Water
Mercury
Mercury
Mercury
Mercury
Mercury
Mercury
Mercury
Mercury

1274 x lOY:;
12~ x lOY:;
127:4: x lOY:;
127:4: x lOY:;
127:4: x lOY:;
127:4: x lOY:;
127:4: x lOY:;
12~ x lOY:;
127:4: x lOY:;
12~ x lOY:;
12~ x lOY:;

Mercury
Mercury
Mercury
Mercury
Mercury
Mercury
Mercury
Mercury
Mercury
Water
Mercury
Mercury

10Ys x llY:;
10Ys x 11Y:;
l0Ys x llY:;
10Ys x 11Y:;
l0Ys x llY:;
l0Ys x 11Y:;
10YsxllY:;
l0Ysx llY:;
10Ys x 11Y:;
l0Ys x llY:;
10Ys x 11Y:;
l0YsxllY:;

Other
structural
variables

Measurements

(1 )
Load
amount,
per cent

(2)
Hook
dist,
in.

Tilt
angle,
degrees

(3)
Lining
thickness,
in.

Deflections
Strains
Side

Bottom

x
x

x
x

x
x
x
x
x
x
x

100
100
25.
50
75
100
100

2Y:;
2Y:;
2Y:;
2Y:;
2Y:;
2Y:;
7Y:;

0
0
0
0
0
0
0

Y:;
Y:;
Y:;
Y:;
Y:;
Y:;
Y:;

x
x

x
x

2Y:;
2Y:;
2Y:;
2Y:;
2Y:;
2Y:;
2Y:;
3Y:;
5Y:;
7Y:;
(4)

0
0
0
0
0
0
0
0
0
0
0

None
Y:;
Y:;
Y:;
Y:;
Y:;
Y:;

x
x
x
x
x
x
x

.....
.....

100
7.2
100
25
50
75
100
75
75
75
91

With both rings

With both rings
With both rings
With both rings
With both rings
With both -rings
With both rings
With both rings
With both 'rings
With both rings
Less lower ring
Less both rings

7.7
50
90
90
90
100
100
98
82
100
100
100

2Y:;
2Y:;
2Y:;
3Y:;
7Y:;
27'2
7Y:;
2Y:;
2Y:;
2Y:;
2Y:;
2Y:;

0
0
0
0
0
0
0
6
20
0'
0
0

Y:;
Y:;
Y:;

With band
Without band
Without band
Without band
Without band
Without band
Without band

.....
.....

.....
.....
.....
.....
.....
.....
.....

Y:;
Y:;
Y:;

Yz

Y:;
Y:;
Y:;
Y:;
Y:;
Y:;
Y:;
Y:;

x
x
x
x
x
x
x
x
x
x
x
x
x
x

x
x
x
x
x
x

x
x

x
x

. x

Loads given in volume per cent of load when liquid level is 3 in. from the lip.
Hook distance is measured from the inside of the shell.
Lining thicknesses given in comparison with unworn lining on actual ladles.
In this test the ladle was supported in stands.

largest on the lip ring and smallest on the lower stiffener,

as are the deflections.
The normal stresses in the side shell are considerably
less than the stiffener ring stresses, except near the
juncture with the bottom plate. The same is the case
for shear stresses in the side shell. Figure 13, giving
. the distribution of vertical normal stresses on ladle "B,"
clearly shows the high local stress peak near the bottom.
These high stresses are similar to the discontinuity
,stresses produced near the heads of pressure vessels.
This stress problem has been' studied by the design
division of the Pressure Vessel Research Committee of
the Welding Research Council, 'and by several others.
Although these stresses sometimes exceed the yield
point, experience from actual ladles seems to show that
they do not endanger the safety of the ladle. The high
discontinuity stresses also occur on the bottom-plate
side of the juncture. Except in this narrow region, the
flat bottoms sustain mainly a high and nearly uniform
bending moment thrgughout, while in, the dished bottom smaller membrane stresses prevail.
The general description of the structural behavior of
the ladle models holds for all tests on all three models,
and constitutes perhaps the most important result of
this investigation. Actual ladles cover a large field of
Reprinted from

IRON AND STEEL ENGINEER, DECEMBER, 1949

types and variations in design. The tests made will

therefore not give the complete picture for any of them,
but the general behavior as described is believed to be
common for all types. The experimental program did,
however, consider some of the more important factors
in ladle design. The variables included may be divided
into two groups, of which the first covers the variation
in testing conditions, such as amount of load, distance
between the hook supports, and angle of tilt. The other
group includes structural variables; riveted versus
welded construction, round versus oval shape, flat or
dished bottom, trunnion assembly size, and type and
number of shell stiffener rings. The effect of these
variables will be discussed in the order mentioned.
The effect of amount of load on the deflections and
stresses is partly demonstrated by Figures 9 to 15, and
so is the effect of variation in the distance between the
two supporting hooks. As a rule, stress and deflection
data for the stiffener rings increase in almost linear
proportion to both load and hook distance, as shown
in Figures 16 and 17, respectively. Since the load distribution varies with amount of load, direct proportionality between load and stresses is not obvious. Table III
shows the critical stresses in the rings for the smallest
and the largest hook distance. A similar, but much less

TABLE III
EXPERIMENTAL STRESSES. PSI
1

---

Locations
0: At trunnions
90: Between
trunnions
o : Outside surface
i: Inside surface

----Lip

Ladle "B"

Ladle "A"
8 x 8 in.
8 x 8 in Trunnions 8 x 16 in.
Trunnions
plus
Trunnions
spacer
band

9
I 10
I
Full load - 7.5 in. hook distance

Ladle "C"

Ladle "A" Ladle "B" Ladle "C"

With both Less lower

No
stiffeners stiffeners stiffeners

Both
stiffeners

8 x 16 in.
Trunnions

-10,800
+28,400

- 8,100
+22,800

- 6,900
+23,400

+ 9,200

- 9,200
+Yield

-11,200
+Yield

-16,700
+Yield

-12,200
+33,600

+17,900

-18,500
+Yield

90

0
i

+ 6,100
-20,900

+ 4,900
-15,500

+ 4,000
-18,000

+ 2,800
- 6,300

+ 3,500
- 5,900

+ 4,700
- 7,700

+10,100
-17,200

+ 7,700
-23,800

+ 5,700
- 7,500

+ 6,000
- 9,500

0
i

- 6,500
+ 3,600

- 5,400
+ 1,200

- 2,200
+ 1,600

-10,000
+ 3,600

- 3,100
+ 3,700

- 3,400
+ 4,300

- 3,200
+ 4,100

- 4,400
+ 3,400

-22,100
+ 6,800

- 6,700
+ 7,300

90

0
i

+ 4,400
- 2,800

+ 3,700
- 1,800

+ 3,500
- 2,000

+ 3,400

+ 5,300
- 3,400

+ 6,700
- 4,600

+ 1,700
- 2,700

+ 5,600
- 3,100

+ 6,400
- 4,200

+ 9,300
- 5,700

- 2,500
+ 2,000

- 1,600
+ 1,900

- 1,000
400
+

+ 1,600
100

0
i
0
i

+ 4,700
- 2,300

+ 4,600
- 1,300

+ 4,200
- 1,900

+ 4,300
- 3,300

+ 4,200
- 2,300

+ 2,000
- 1,300

+ 3,600
- 2,700

+ 7,600
- 4,200

+ 8,400
- 7,000

+ 4,400
- 4,400

280

350

310

250

340

333

325

310

250

310

----

----

stiffener

---- - - -- -----

stiffener
90

----Ladle dead weight,

without lining, Ib

Full Load
l_evel-----L.J

II

II

2.5 or 7.5
Hook Dis tonee

consistent variation is obtained for the side shell. Bottom stresses and deflections also increase with the load,
but less rapidly than in the rings. This should be
expected, because membrane stresses prevail in the
bottom when the deflections become larger than the
plate thickness. Bottom stresses and deflections show
very little response to an increa e of the hook distance.
One test was made with a ladle supported in stands,
resting on the underside of the trunnion assemblies.
This condition is essentially equivalent to a smaller
hook distance, and the re ults fall in line with those
given above.
Ladle "C" was tested in tilted positions up to 20
degrees, with no increase of the stresses. At more than
20 degrees, the stresses started to decrease due to outpour of the liquid load.
Ladle model "A" is riveted, while the two others
represent welded construction. The three ladles differ
also in other ways, and the test results are therefore not
directly comparable. As an average, under equivalent
conditions, the magnitude of the side deflections are
about thrce times as large for the riveted ladle as for
either of the two welded ones. This is at least partly
explained by the fact that the average ratios of the
bending stiffness of the stiffener rings above and below

0
i

- - - ---

Lower

4
5
I
I
I
Full load - 2.5 in. hook distance
3

ring

Top

400
100

100
100

--------

----- ---- ----- 1,300 + 3,100 200

700 +
300
300 +
-------

100
100

the trunnions are one third as grcat in the riveted ladle

a in either of the welded ladle. The experimental
tres es, essentials of which are given in Table III,
show no consistent difference between riveted and
welded construction. The structural efficiency of the
ladles should be compared on a dead weight basis, since
maximum live load within the ladle crane capacity
limit is of prime intere t. These weights are therefore
given on the bottom of Table III.
Ladle "B" is an oval ladle, while "A" and "C" have
circular cross sections. The deflections show the same
trend for all three ladles, and the stresses do not allow
any conclusions as to preference. It is believed that the
mall degree of ellipticity commonly used has but a
minor influence upon the structural behavior.

Figure 4 -

Repl'inted from

Trunnion assembly for ladle "A."

IRON AND STEEL ENGINEER, DECEMBER,

1949

A study of the results from the tests of the flat

bottoms on ladles "A" and "B" and the dished boLtom
on ladle "C" allows definite conclusions in favor of the
dished type. As mentioned earlier, the dished bottom
plate carries the load mainly through membrane
stresses. The resulting deflection, Figure 15, is therefore
only one sixth to one scventh of those dcvcloped in the
flat bottoms, as shown in Figure 10. The figures also
show a different deflected shape for the dished bottom,
which does not have the maximum deflection at the
center point. The only re emblance between the two
types is the in ensibility to variation in hook di tance.
The flat bottom plate ustain high bending tre es,
while the di hed hape show tensile stre e throughout
its thickne of about one fourth of the magnitude of
the measured stresses on the flat bottom . The discontinuity stresses both in the side shell and in the bottom
plate near the knuckle is even more reduced when using
a dished bottom.
Ladle "A," the riveted model, wa te ted with two
sizes of trunnion assemblies, 8 and 16 in. wide, covering
34 and 68 degrees of the ladle circumference, respectively. The effect of the wider trunnions is to decrease
the ring stresses, as seen from columns 1 and 3 of
Table III. The reduction is largc t on the lip ring and

Figure 5 -

r
r
I
I

Trunnion assembly for ladle "B."

on the top stiffener near the trunnion, and averages

43 per cent. The side deflections with the wide trunnions are about 75 per cent of those obtained with the
narrow type. The bottom stresses and deflections stay
practically unchanged.
Columns 1 and fl in Table III give information on the
effect of the addition of a spacer band on ladle "A."
This >i X 8>i in. band, which is shown in Figure 1, is
riveted to the sides of the trunnions, but not connected
with the ladle hell. It is intended to carry most of the
bending moment from trunnion to trunnion, thus
relieving the regular reinforcing rings. Tablc 1I I shows
an avcrage reduction in ring stresses with 26 pel' cent
of those obtained without the spacer band. The additional weight of ladle steel material due to the band is
fl5 per cent. Stresses in the side shell are also decreased,
although not altogether consistently. The bottom
stresses are not affected by the addition of the spacer
band. A comparison of columns fl and 5 in Table III
seems to indicatc that the same advantage may be
Reprinted from

IRON AND STEEL ENGINEER, DECEMBER, 1949

Figure 6 -

Trunnion assembly for ladle "C."

obtained in a cheaper way by adding the equivalent

steel weight to the ladle shell thickness between the top
and lower stiffeners. Higher lip ring stress is registered
in the latter case. Deflection tests show, on the other
hand, that the heavier shell more effectively resists the
local bending effects caused by the moments introduced
through the trunnions.
Finally, the effect of removal of stiffener rings was
investigated on ladle "C." Some of the test stresses are
recorded in columns 5 to 7 in Table III. It shouldlbe
remembered in interpreting these results that ladle "c"
has a heavy middle shell section and correspondingly
wcak rings. Removal of the lower stiffener ring increa e the stresses on the top stiffener by 23 per cent
and the lip ring stress by 27 per cent on an average.
With both the top and lower stiffeners removed, the
lip ring tresses increase to an average of 143 per cent
above those recorded with both stiffeners in place. No
test has been made with only the top stiffener removed,
but the tests made indicate that the top stiffener is
comparatively more efficient than the lower stiffener.
The deflection tests, Figure 15, also point to this conclusion, since the deformation at the top ring is considerably larger than at the other rings.

DESIGN RECOMMENDATIONS AND STRESS ANALYSIS

The conclusions which can be drawn from the test

results, as presented in the previous chapter, are incorporated in the design recommendations below, which
also summarize recommendations made in various
articles on ladle design. Statements which are referred
to articles and not to parts of this investigation should
be regarded as the opinion of the author of the reference.
M alerial - Ordinary low-carbon steel is recommended for hot-metal ladle construction. The temperature
effect and high cost di courage the use of high-strength
alloy as a means of reducing the dead ,,,eight. For the
trunnion, a 3 per cent nickel, low carbon steel is
sometimes advised as being stronger and more resistant
to pills of metal and slag (4). Reference 6 recommends
forged steel ASTM A-fl35 Class B for the trunnions.
For trunnion pins and other parts for which reliable
design methods allow the use of high stresses, the im-

portance of a smooth finish of the entire surface is

emphasized (3).
Main ladle p1"oportions - The ladles are usually
shaped like a frustum of a cone with a side slope, or
increase in radius, of one inch per 12 to 15 inches of the
height.
The quasi-elliptical or oval cross section shape,
introduced to increase the capacity under existing mill
conditions, is usually obtained by inserting a straight
shell section at the trunnions. By making these middle
sections slightly curved, lining conditions are improved
(5). The ratio of minimum to maximum cross section
width usually falls between 1 and %. On the oval ladle
model this ratio is 0.82. No noticeable difference was
found in structural behavior between the circular and
the oval ladle models. The round shape is preferred
because of considerations of manufacturing cost and
lining conditions.
In Figure 18, the ladle height and diameter, or
average diameter for oval shapes, is given as a function
of the carrying capacity for the ladles listed in Table I.
A similar diagram for capacities up to 100 tons is given
in Reference (3). It is seen that the height is slightly
larger than the diameter up to approximately 12-foot
height, or 125-ton capacity. Above that tonnage, the
height is smaller than the diameter. In Reference (4)
it is advised that the diameter only should be increased
above 125-ton capacity in order to keep down the head
of metal during pouring.
The location of the center of gravity of a fully loaded
ladle of 12-foot height should be about 15 inches below
the trunnion pin center line, and correspondingly placed
Figure 7 - The general location of the gages are shown in
this drawing for ladle "A." Similar locations were
used on the other ladles.

Figure 8 - In this setup for the strain tests, the equipment for measuring the strains is seen in the foreground. The ladle is seen in the background in the
vertical position. Tape seen on the ladle is used to
keep the leads from the electric strain gages in place.

for othel' size ladles (4). A method for calculating the

location of the center of gravity and the tipping moment
is given in Reference (10). In Reference (6) it is recom0.00 C
mended that the center of gravity be d=6
jnches below the trunnion center line, where C is the
ladle capacity in net tons. If this formula is used, lock
bars to prevent tilting should not be required.
Type of construction - The two commonly used types
of ladle construction, riveted and electric arc welded,
were both represented in the experimental investigation.
The riveted model underwent much larger deflections
than the welded ones under equivalent conditions, but
this was at least partly due to the lighter design of the
riveted model. The experimental stresses showed no
consistent difference between the two types.
The steel weight of the riveted ladle model was 280 Ib
as compared to 250 and 340 Ib for the two welded models
(Table III). In actual ladle design, the older riveted
types, without spouts, lining and stopper rings, weigh
21 to 27 per cent of the rated capacity. The same percentage for similar welded typcs of more than 50-ton
capacity is only 16 to 17 per cent. Welding gives
lighter ladles for capacities above 60-70 tons. Considering the weight of slag, lining, etc., welded construction
allows an increase in capacity above that of riveted
construction by about 7 per cent for a 150-ton ladle (4).
The joints in the shell should be butt welded. All
welding should be specified in accordance with the

0-1-1-+-.

Identical \lOges on inside and

autside except an rings.
AI

, Mechanical
-

2
3

5
6

7
8

10

Gages

SR-4 Gages

Reprinted from

IRON AND STEEL ENGINEER, DECEMBER,

1949

DEFLECTION

HCAlZOHTAL

LIP RING

CROSS

SECTION

AT:

TOP STIFFENER

---,

6150-1.50
'150"2.50
}115"Z.50

with full liquid load and the longest distance between

the supporting hooks. Tilting of the originally fully
loaded ladle has no increasing effect upon the stresses.
It follows that, in practical design, the ladle should be
assumed loaded to the slag spout level with liquid metal,
and the hook spacing should be chosen no larger than
required by operating conditions.
Dead weight assumptions - The specific weight of
molten metal varies with the grade of the steel. A suitable design value is 420 lb per eu ft. As mentioned
earlier, the net dead weight of the ladle is close to 74
of the rated capacity for riveted ladles, and 1/6 for
welded types of more than 50-ton capacity. Spout,
stopper, etc, will increase these fractions by 10-15 per
cent. The weight of the fire-brick lining will vary with
the preferred thickness. Common lining thicknesses are
9-14 in. on the bottom, and 9 in. decreasing to 7 in. on
the sides from the bottom up. For ladles of about
150-ton net capacity, the lining weight is ordinarily
80-85 per cent of the total steel dead weight.
Trunnions - The trunnions on ladles of 100-ton
capacity or more are usually of welded construction.
Castings are now mostly used for smaller ladles. The
trunnions should be connected to the stiffener rings
which are usually provided above and below the trunnions in order to minimize loeal shell stresses due to the
trunnion reaction. Tests show a beneficial effect of
comparatively wide trunnions.
Figure 18 shows recommended trunnion pin diameters
for different ladle capacities. Reference (6) advises that
the pin diameter should not be reduced where it fits
into the trunnion plate, and if the pin diameter is
increased at that location, a large fillet radius should
be provided. The stresses in the trunnion pins can be
calculated by means of conventional methods.

\,

LOWER STHENER

~.~i~m:~

63~o-150

63&0"-%'0
51U"Z-'O

s.cel. r.. O.II",,_

~

Figure 9 - The deflections for ladle "A" through horizontal planes are given for various loads and for two
positions of the ladle hook on the trunnion.
Figure 10 - The deflections for ladle "A" through vertical
planes are given for various loads and for two positions
of the ladle hook on the trunnion.
A

seD
n.

cm..ECT()NS

OF

\IEllTJCAL

-_.,,~

-m
:~

-3i7s- -

,~

,-

1115

MlIollObl... f"taIlhlid!t

'" 6Y!4.7.51)

1t"64-2~

c- 41"'.2.'0"
f,I1t'.UCf

,. tltd'.uo-

applicable codes of the American Welding Society. The

opinions on the necessity of stress relieving after welding
seem to be divided. Most manufacturers, it is believed,
stress relieve the ladles at 1200-1300 F for one hour
per inch thickness of the heaviest material.
Design assumptions - The design assumptions should
be in agreement with the general structural behavior
observed during the tests. The ladle sides are found to
deflect inward in their full height in the trunnion
region, and outward in the region midway between the
trunnions. The deflections are nearly zero at the bottom
juncture, increasing approximately linearly towards the
maximum deflections at the lip ring. A circular ladle
cross section will, accordingly, tend to become elliptical
under load, the amount of ellipticity increasing from
bottom to lip. Vertical lines on the side shell located
about 45 degrees from the trunnions experience no
deflections. Thus, the maximum side deflections at each
level occur along vertical lines through the trunnions
and half-way between the trunnions, the absolute
maximum being obtained at these locations on the lip
rmg.
The highest stresses on the reinforcing rings were
obtained at the same locations, and followed the same
pattern as the maximum deflections. High stresses were
also measured in the bottom plates, and discontinuity
peak stresses were found in a narrow region on both
sides of the side-bottom juncture.
At these critical locations, as well as on other parts
of the ladle, the most unfavorable condition is obtained
Reprinted from

IRON AND STEEL ENGINEER, DECEMBER,

~~lood.21_~

C&:I"Lood.21" H_ Olttonc.

C<

LaM.

7( __ DiIlo_

Figure 11 - The horizontal stresses on the outside of the

ladle "A" are given for several loadings and hook
distances for the ladle in the vertical position.
Figure 12 - The horizontal stresses on the inside of the
ladle "A" are given for several loadings and hook
distances for the ladle in the vertical position.

Scol,

rOf"

"n L.Ht. 21"

.. ~~2~

..,.,

SlrtsSlI
0i.1OII<'

-..

c,.".;'~_rl~-...

1949

11

1.-

A
~

...

j/

../

. .-=- . .

-.-... ~

_. Figure 13 - Outside stresses through vertical planes of

ladle "8" for two loads with a hook distance of 2Y2 in.
from the inside of the shell.

Side shell - The thicknesses commonly used for ladle

side shells are within 10 per cent either way from those
shown in Figure 18. In order to obtain comparable
stresses in the model and prototype, the model shell
thicknesses were made as small as possible, considering
available materials. The model corresponds to a % in.
thickness on the 150-ton capacity prototype. The
stresses in the model, loaded with mercury, should be
about 3/5 of the stress in the corresponding prototype,
loaded with molten metal. No severe stresses were
observed, except near the bottom juncture, as shown in
Figure 19, which records the average of the outside and
inside side shell and bottom stresses for a section
through ladle "C." Practical operating considerations
probably require a thicker ladle shell than required by
stress analysis, and the former will therefore be the
determining factor in selection of the shell thickness, at
least for lfidles up to 150-ton capacity.
The problem of a strict analytical determination of
the discontinuity stresses in the side shell near the
bottom juncture is a very complicated one in the case
of a ladle. The trunnion reaction caused a non-uniform
distribution of these stresses around the ladle circumference, with a peak directly below the trunnions. Even
on flat bottoms, a small knuckle radius is ordinarily
provided, which further complicates the analysis. On
riveted ladles, the lap-joint and the discontinuous connection add to the inadequacy of conventional stress
calculation methods. Most disturbing, however, is the
effect of the very large deflections of flat ladle bottoms.
A complete theoretical attack on the problem (not
yet finished for use in practical design) is given in
Reference (ll). The problem is treated in a more
practical way in Reference (12), and also in Timoshenko's "Theory of Plates and Shells," Chapter XI.
The procedure below is based on the two latter references. Development of the formulas may be found in
Reference (13).
With notation as explained in the nomenclature, and
disregarding all the disturbing factors mentioned above,
an impression of the magnitude of the discontinuity
stresses may be obtained by the following procedure.
Let:
3(I-v 2 )

2.73

----'-------'- = .- . . . . . . . . . . . . . . . . . . . . . . . . . . . (1)
r 2t 2
r 2t 2

12

Figure 14 - The deflections through horizontal planes of

ladle "e" are given for no-tilt of the ladle.

The moment per inch of the circumference acting on

the bottom end of the side shell is then, with Poisson's
ratio taken as 0.3:
3
Mo = (-pr) 0.525 rt 0.425/ h
(2a)
4.200 t + (33 3r

A positive moment 'caused tension at the outside shell

surface. The radial end shear at the same location,
positive when acting inward on the side shell, is:
Po

0.4:5 p -

(3Mo

(3a)

The moment acting in the vertical direction per inch

width of tlie shell at any point at a distance y from the
bottom juncture is:

+ ~o ~

My=Mo c/J

(4)

and in the circumferential, or horizontal, direction:

M x = 0.3 My
(5)
The functions
c/J=e - {3y (cos (3y
sin y
~=e
sin (3y
if;=e - / (cos (3y - sin (3y)
(j = e - Y cos (3y

-:y

where e is the base of the natural logarithms, are

tabulated for values of (3y between zero and 7.0 on
page 394 of Timoshenko's "Theory of Plates and
Shells," First Edition. Considering both the ordinary
membrane stresses and the discontinuity effect, the
total stresses in the shell near the flat bottom are then:
pr
6My
+2t - t 2

(7)

CTy= -

in the longitudinal direction, and:

CTx=P: -

~r

[(3Mo if;

+ Po (j]

6~x ..... (8)

in the circumferential direction.

From the table referred to above it may be seen that
the discontinuity stresses decrease rapidly with increasing distance from the bottom juncture. Beyond (3y = 1T',
or, using Equation (1) and v = 0.3,
y=2.44
Reprinted from

vrt
IRON AND STEEL ENGINEER, DECEMBER,

(9)
1949

TABLE IV
COMPARISON OF EXPERIMENTAL AND COMPUTED NORMAL STRESSES IN THE SIDE SHELL NEAR
THE BOTTOM JUNCTURE MIDWAY BETWEEN THE TRUNNIONS- FULL MERCURY LOAD
<TX:
<Ty:

Distance
from
juncture
(in.)

Ladle

horizontal
vertical

outside
i: Inside

0:

Ratio:
computed
experimental

Computed
stresses
(psi)

Experimental
stresses
(psi)

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - --------- --------+1,600

"A"

3.43
Flat bottom

800

0.50

1.00

------- --------

(Jx

- - - - - - - ---------+1,800

0.82

0.54

+2,200

-1,300

-8,100

-8,200

-5,600

-6,000

-4,~00

------- - - - - - - - - - -

<Ty

------- ------- --------

"B

700

- - - - - - - - ----------1.01

----------

<Tx

--------

1.43
Flat bottom

"e"

<TX

Dished bottom
(d =1.90)

0.67

+4,400

+4,100

0.93

-8,200

3.28

- - - - - - - - - - - - - - - - - -----------

-2,500

-4,500

+2,100

-1,200

-------- --------

0.10

-3,100

-------- ----------

<Ty

-------- -------- -------

1.07

----------

- - - - - - - - - ----------2.04

-9,200

----------

+2,000

0.95

-1,100

0.92

--------- ---------

<Ty

These stresses are without any practical concern. The

calculation above, on the other hand, will also show
that the maximum stress for ordinary ladle dimensions
exceeds the yield point. The maximum stress will, in
most cases, occur in the longitudinal direction for y = 0,
and can then be computed by putting My =Mo in
Equation (7).
It can be shown, that the most favorable bottom
shape from a combined economy and stress analysis
point of view is an ellipsoid with a depth of about one
quarter of the bottom radius. However, a shallower
bottom, as generally used for ladles, in comparison with
a flat bottom will greatly reduce the discontinuity
stresses. The stress calculation in case of a dished bottom
follows the same procedure and is subject to the same
reservations outlined above for flat heads. Only the
expressions (2a) and (3a) will be changed. Assuming
that the ordinary two-radius dished bottom deviates

little from the ellipsoidal shape, and that the knuckle

radius is comparatively large, these expressions are
substituted by

Figure 15 - The deflections through vertical planes of

ladle "e" are given for no-tilt of the ladle.

Figure 16 - The measured ring stress for different weights

in the ladle indicate that they increase directly with
the load.

DEFLECTIONS Of

VERTICAL CROSS S[CTIONS AT A. B, Co 81 0 ~

A B C
3450" 6160"

~- 6860"1810'"

68fi0"

I'

7'

2 '21"1'

MIO'~ 34Jd"

7"1."21'

1-

.-f-I--I

:= _
4-

6810'

6860"

11"

zi"

~~

- -

rr

Mo=( -Po)
P _ (
0-

(2~~~: ~ ~1:;hO'5)
2

p)(

2~d2h 4+

rt
1.7 (h-t)d
3h 2 - t 2 U - tI.5
h2
t2
hl.5

[h

.. (3b)

pr 2

Po = 813 d 2 '

(Sc)

The tests of the ladles were not specifically designed

JtF

LADLE

LAOLE "A"

00.

cr0

."
'"

..
~

unr
ito

." iimo,

L
I
-'-'!"

'c'

U,

)~~ti\'_.

II " "

For bottom thickness h equal to the side shell thickness

t the expressions become
M o =0
(2c)

~:~

+
+

....... :.(2b)

."......
...
'"

eli

1-'

!-.

.,

-,

I,
-1'18.16

."

".
Reprinted from

IRON AND STEEL ENGINEER, DECEMBER,

1949

13

LADLE "B"
8

10

3.5

7.5

~
d 6
en

A3i
010
C30

o
en

:3 -2
II:

:n -4

~~~~::;J+BI 0

Oiatanci

B50

I:)

~-6
-8

CI i

LADLE C
~ 8

II:

~ 4

lit
~

2.5

3.5

030 0: Outside Gage

7.5

.
A3\
010
C 30
Clo

i: Inside Gage
A5,0I,C3. etc indiccile llaQ41 10calion.lI. Fill. 7

LJ-------<r C 50

o
C5i

F::;:::::::-----lB I 0
C3i
051

-8

A30

Figure, 17 - The ring stresses increased directly with the

moment arm of the ladle hook on the trunnion.

to investigate the problem of, discontinuity stresses,

which in themselves constitutes a large field of research.
The deviations of a ladle from the ideal cases covered
by the above formulas are also too many and significant
to allow a close check with experimental stresses. A
comparison, Table IV, is therefore carried out only for
one gage point on each ladle located near the bottom
juncture midway between the trunnions where the disturbances are smallest. The discrepancy in Table IV
for the horizontal stress on ladle "C" may be explained
by the location of the gage, which was mounted very
close to the reinforced bottom butt weld. Also, the
expressions (2b) and (2c) are not exact.
The corresponding experimental discontinuity stresses
directly below the trunnions are two to four times as
large as those midway between the trunnions. The calculated maximum stress at the juncture (y = 0) is far
above the yield point for ladles "A" and "E" with the
flat bottoms, and 10,300 psi for ladle "C" with a dished
bottom.
Bottom plate - A dished bottom is recommended as
advantageous in several ways as compared to the flat
type. The dished bottom is usually of the shallow dish
type tank head, shaped like a spherical segment, with
a smaller knuckle radius giving an arch tangent to the
head flange (sometimes called torisphericaJ). The maximum deflections of tl).e flat plates were six to seven
times as large as the maximum measured for the dished
bottom, and exceeded the thickness of the bottom

14

including the cover plate. The maximum deflection of

the dished bottom was only about 0.2 of its thickl).ess,
and the latter also gave smaller over-all stresses. The
discontinuity peak stresses at the juncture of side and
bottom are greatly reduced when a dished bottom is
used, especially if a large knuckle radius is provided.
Flat bottoms usually become semi-dished after long
service, indicating that the yield point is exceeded in
some parts of the plate. The dished type provides
inter-locking of the bottom fire-bricks, decreasing the
danger of damage due to the liquid flotation force. One
practical advantage of using a flat bottom is that the
ladle may be set directly on the ground without danger
of tilting.
.
The bottom should be furnished with a removable
cover plate for protection against heat radiation during
the pouring process. The cover plate is spaced away
from the bottom by means of washers, and can be
renewed or removed to allow inspection. The cover
plate extends over the whole bottom or parts of it,
according to the pouring schedule applied (4). A similar
plate is also recommended underneath the slag spouts.
extending down to the lower stiffener ring (6). Pouring
openings in the bottom should be reinforced like holes
on pressure vessels.
Common thicknesses of the main bottom plate, flat
or dished, are shown in Figure 18. The thickness is
usually a little larger than for the side shell, but may
vary up to 20 per cent in either direction from those
indicated in the figure. The tests showed that the
bottom stresses depend only upon the depth of the
liquid load, and that an analytical determination of the
plate thickness mU3t include both bending and membrane stresses.
'
For flat bottom plates, such a stress computation
may be based upon the diagrams in Figure 20, reproduced from Timoshenko's "Theory of Plates and
Shells," First Edition, pages 340 and 341. A ladle bottom edge is not completely fixed, as assumed in the
diagrams. The strain measurements on the bottom
were few (Figure 7), and do not coincide with the
location of the stresses given in Figure 20. Although
disregarded in actual design, the presence of the cover
plate on the ladle model impedes a verification of the
computed stresses. Using the added thickness of the
main plate and cover, the diagram gives a maximum
deflection of 0.17 in. as compared to 0.21 in. measured.
The calculated model stresses at the center for full load
are +22,100 psi at the outside and -11,700 psi at the
insid~ surface. The corresponding experimental 'stresses
at 0.32 times the bottom radius from the center point
were +18,900 psi and -14,60'0 psi. The plate connection efficiency factor and the smaller actual degree of
edge restraint add to giv~ a too low calculated deflection
and bending stress, and subtract in their effect on the
calculated membrane stress. The maximum computed
stress at center, which occurs on the outside surface, is
believed to be a good criterion of the required plate
thickness. The edge stresses will be smaller than given
by this method due to imperfect clamping.
For a ladle of 150-ton capacity the empirical diagram
in Figure 18 indicates a 1.3 in. thick bottom plate.
With other dimensions as given in Figure 18, the stress
in the bottom plate is found to be +33,200 psi and

Reprinted from

IRON AND STEEL ENGINEER. DECEMBER. 1949

-18,000 psi at the outside and inside surface at center,

respectively. The experience that flat bottoms will
undergo permanent deformation during service thus
seems verified by stress calculation. Whether yielding
should be tolerated as in the past, or be prevented by
increasing the thickness or using a dished plate, will be
a matter of judgment.
ASME's Boiler Construction Code of 1946, Sections I
and VIII, for Power Boilers and Unfired Pressure
Vessels, and also the joint API-ASME 1938 Code for
Unfired Pressure Vessels for Petroleum Liquids and
Gases, require a minimum flat head thickness h inches
according to the formula
h=dl l c:

(10)

where
d=plate diameter, or shortest span, in.
p = maximum load, psi
s = allowable unit working stress, psi
c = 0.25 for butt welded circumferential bottom joint
c=0.30 for lap-riveted circumferential bottom joint
For the 150-ton prototype with 1.3 in. bottom plate,
this formula gives 16,000 psi maximum stress. The
formula thus leads to thicknesses of the order now
ordinarily used, but fails to warn against the actual
high stresses.
The codes mentioned above also give a formula for
the required thickness of dished heads:
5 pR
h=- 6 s

(11)

where the symbols have the meaning as explained for

Equation (10). However, the codes restrict the radius of
curvature, R, to be smaller than the bottom diameter.
For hot-metal ladles this radius is usually much larger,
and a too small thickness therefore results from this

Figure 18 - The variation of ladle dimensions for various

capacity ladles is given in this diagram.
AVERAGE

-0-

-1-

~
J:

DIMENSIONS BASED ON TABLE

OUTSIDE HEIGHT
OUTSIDE TOP DIAMETER
100 x BOTTOM PLATE THICKNESS
100x SIDE SHELL THICKNESS t

20~

I O~

120

~
~

80

If)

en
z

ILl

2:

is

20
10
0

40
RATED

Reprinted from

80

120

CAPACITY

IN

160
TONS

200

formula. Assuming radii of bottom curvature propor-'

tiona I to the one used on the model (Figure 3), Equation
(11) gives a maximum bottom stress of about V5,,000 psi
for ladles of ordinary dimensions as shown in Figure 18.
The actual stresses are larger. Thus, for the dished
bottom model, Equation (11) gives Smax = 3,700 psi,
where 8,100 psi was measured near the bottom center.
The discontinuity stresses near the bottom edge are
of a magnitude similar to those earlier discussed 'for the
side shell. These stresses can only be reduced to below
the yield point by using a dished bottom, and therefore
were not discussed in connection with the design of flat
bottoms.
A stress analysis procedure for dished bottoms will
not be attempted here. The problem is discussed in the
article "Stresses in Dished Heads of Pressure Vessels"
by C. O. Rhys, ASME Trans. 1931. When finished,
the work of reference (11) will probably give the most
complete answer.
Reinforcing rings - The ladle models all had stiffener
rings at the lip and above and below the trunnions.
Investigation of the effect of the two trunnion rings
indicated that the most efficient use of ring material is
obtained at the location above the trunnions. Tests
alone would suggest emphasizing the lip ring in ladle
design. Practical experience shows, however, that the
lip ring often suffers great damage from the hot slag
and metal splash during service. The safety of a ladle
should therefore not be made dependent on the presence
of this ring. Also, the actual load carrying contribution
of the lip ring may be lessened because of the discontinuity introduced by the spouts, which were eliminated
on the models. It is therefore not recommended to follow
the test indication towards an increase of lip ring dimensions. In order to meet the condition after long time
service, the stresses should be calculated both considering and disregarding the lip ring.
Reinforcing of the shell against bending and twisting
moments is sometimes "accomplished by other means
than the stiffener rings discussed above. A wide stiffener
band, connected to the trunnions but not to the shell,
reduced the experimental stresses. However, a similar
result seems obtainable by adding the equivalent steel
material to the regular stiffener rings or to the shell
thickness between the trunnion rings. Reinforcing by
means of a heavy middle shell section and rather small
rings, as on one of the ladle models, was found to give
a very rigid ladle with small deflections, but the stresses
were not correspondingly reduced.
The locations and conditions for the maximum
stiffener ring stresses are discussed earlier. It follows
that, in designing the rings, it is sufficient to calculate
the stresses at the trunnion line and midway between
the trunnions on each ring for full liquid load and the
largest hook spacing which may occur during operation.
The adjacent rings above and below the trunnion
have an increased stiffness by virtue of their connection
to the trunnion assemblies. Even if the trunnion assemblies are connected to the side shell only, as in the
case of ladle "A" (Figure I), the rings will be somewhat
restrained in the neighborhood of the trunnions. Due
to this effect, the two trunnion rings will ordinarily
obtain the maximum stress midway between the trunnions, as may be seen from Table III. The increase in
ring stiffness at the trunnions is not easily calculated.

IRON AND STEEL ENGINEER, DECEMBER, 1949

15

Section 0-0

fff
o

side shell, this method gives u~satisfactory numerical

results for hot-metal ladles.
The method given below is partly based upon general
trends of the experimental results, which limits the
range of its usefulness to ladles of types not basically
different from those tested. Details in development of
the procedure are avoided here, and are given in
Reference (13).
The following are the assumptions on which the stress
analysis procedure is based:
1. The deflections of the ladle side increase linearly,
from zero at the bottom juncture, to a maximum at
the lip.
~. The rings adjacent to the trunnion assemblies are
subjected to concentrated inward transverse loads,
as shown in Figure ~1. When the lip ring is disregarded these loads are the only ring loads, and
must therefore, acting separately, cause deflections
in accordance with assumption 1.
3. The lip ring deflects in agreement with assumption 1
due to a loading imposed by a side shell shear flow q
psi which, on a horizontal section through the ladle,
is distributed according to
q = a2 sin '24>
a4 sin 44> ................. , .. (12)
where a2 and a4 are different in each of the shell fields
between two rings or between the lower stiffener and
the bottom. The angle 4> is measured as shown in
Figure ~3.
4. The stiffener rings would .experience no stress if the
ladle were supported on an inward extension of the
trunnion pins at points determined experimentally
(Figure ~3). The first assumption is in agreement
with the test experience, demonstrated in Figures 10
and 15, that the ladle side deflects inward at all
levels in the trunnion region. It was sometimes
assumed in ladle design that the lower stiffener
deflected outward in that region due to the moment
introduced through the trunnion pin. The tests
showed that this effect is of negligible importance.
The two concentrated loads P on each ring at
each trunnion represent a simplification of the actual
distribution of the load imposed upon the ring by
the shear flow in the side shell. Assuming a triangular
resultant load over the width' of the trunnion
assemblies, Figure ~~, and P-Ioads are one third of
the total trunnion width apart. This assumption
gives stresses for different trunnion widths which are
in good agreement with the experimental results.
The. use of concentrated loads on the rings adjacent to the trunnion assemblies, and a distributed
loading on the lip ring, is most readily explained in
connection with the fourth assumption. The radial
deflections of a circular ring of uniform moment of
inertia subjected to concentrated loads as explained
are

4 5
Stress SCale
Kips/In"

6 7
8 9

Figure 19 - Graphical view of horizontal membrane

stresses on a vertical cross section halfway between the
trunnions is here given for ladle "e" for full load,
with no tilt of the ladle, and with a hook distance of
2Y2 in.

If the degree of restraint is obviously' large, as for

ladle "C" (Figure 3), the problem is of no consequence
since the stresses midway between the trunnions will
govern. The same is the case when wide trunnions are
used: Rows 1 and 3 of Table III show that an increase
of trunnion width from 34 to 68 degrees of the ladle
circumference moves the critical section of the top
stiffener from the trunnion line to midway between the
trunnions. Without the increased strength due to the
trunnions, all three rings would obtain maximum stress
at the trunnion line. As earlier recommended, the
,trunnions should therefore be made comparatively wide,
and be conneCted directly to the stiffener rings. Following this recommendation, it is necessary and sufficient
to calculate the stresses at the following locations:
Lip ring (Ring 1): 0 and 90 degrees
Top stiffener (Ring ~) : 90 degrees
Lower stiffener (Ring 3) : 90 degrees
The designations are explained on Table III.
An exact analysis of the stresses in a hot-metal ladle,
if possible, would be extremely complicated. A practical
procedure must compromise between exactness and
simplicity. The semi-rational method proposed emphasizes the latter. A more theoretical attack was
attempted, using methods outlined in the article
"Stresses in a Reinforced Monocoque Cylinder Under
Concentrated Symmetric Transverse Loads," by N. J.
Hoff, Journal of Applied Mechanics, Vol. 11, No.4.
Probably due to neglecting the bending stiffness of the

16

[7f'

pr 2'-a +( ~-cosa).
.
)]
Ilr = ~EI
sma-:;4 ( asma+cosa
.
at the trunnions (0 degrees), and
3

pr [:;
4(
.
Il r , = ~FI
asma
+cosa) -1
midway between the trunnions (90 degrees), an
increase in radius being taken as positive. The foreReprinted from

IRON AND STEEL ENGINEER, DECEMBER,

1949

going equations may be obtained from Reference

(14), page 153. Assumption 2 then leads to the
following relation between the loads P 2 on the top
stiffener and P a on the lower stiffener:

"

Ba

P a =B P 2 (13)
2

where B is the product of the ring bending stiffness and

the distance of the ring from the bottom. For the lip
rmg,

B1=EI 1 h.
rIB

and for the top and lower stiffener the subscripts 2 and
3, respectively, are used.
The skin shear is set up by the tendency for relative
tangential movement between a ring and the adjacent
side shell when the ring deflects. 'The skin shear must
therefore be zero at cP = 0 degrees and cP = 90 degrees
(Figure 23), where the ring deflections are purely
radial. The shear distribution must also be symmetrical
about the two axes of symmetry of a horizontal ladle
cross section. A sine-series, using only the two first
terms as in Equation (12), will satisfy these requirements. The additional terms cause negligible ring
stresses, and are dropped.
The resultant skin shear load on a ring equals the
difference in shear in the two shell fields adjacent to the
ring. Writing
C22 = r2a22 - rla12.'
(14)
etc., where the first subscript refers to the ring number
and the second subscript to the sine series term number,
Figure 20 - The deflections and stresses in a uniformly
loaded circular plate with a clamped edge is given in
this figure as presented by Timoshenko's "Theory of
Plates and Shells."
1.2
JI. 0.3
I.'"

In ..

V-

kl,6
0.

0,2/

.-.

----

Positive directions os indicoted

Figure 21 - The free body diagram which illustrates forces

acting on the ladle are given in this figure.

the loadings over an arc d~ of the circumference of the

rings are, avoiding details in the development,
qlrld~=cI2 sin 2~
C14 sin 4~
,)
q2r2d~ = C22 s!n 2~ + C24 s!n 4~
(15)
qarad~=ca2 Sill 2~
Ca4 Sill 4~
For simplicity, these loadings are assumed to act at the
neutral axis of the rings. (Figure 21.)
The fourth assumption listed above is in fair agreement with the deflection pictures, Figures 10 and 15.
With the assumed P-Ioads and shear loadings, Equation
(15), the equations expressing straight-line deflection
of the sides at the trunnion line and midway between
the trunnions may be written (13):

+
+

C22=:: C12 - P 2 g (a)

/'

7'

.................. (16)
2

LOAD.

pr 4

10

12

E"'hi'

3--,.----.-------.----r----.-----,

Ca4=::C14 - Paf (a)

where g (a) and f (a) are given by Equations (24) and

(25).
en
en

I&l

II:

Iii

0.4

0.6

DEFLECTION

Reprinted from

W~

0.8

LO

1.2

IRON AND STEEL ENGINEER, DECEMBER,

Figure 16 shows that the ring stresses are nearly

proportional to the amount of load. Figure 17 indicates
. a linear relationship between the ring stresses and the
hook distance. If the stress diagrams in Figure 17 are
extended to the left, they will all cut the horizontal
axis (T = 0 at or ne'ar one point, located approximately
a distance rt - r sin 45 degrees inside the ladle shell.
1949

17

'Trunnion

(4wa)(

+ B12h 2 + Bah a)
(23)
g(a) =4..5 [0.5708 - a + (2-sin a - cos a) sin a]
.................... (24)
f(a) =225 [2.5708 - a + (2 + sin a - cos a)sin a
k=

P--If"~

BIh

- 2.5464 (a sin a + cos

to determine the ring loadings:

P _
2-

Stiffener Ring

)]

(25)

(wa)
(
4

)
B2
B 2h 2 + Bah a

B
B2

P a= -a P 2

Figure 22. - A triangular resultant load is assumed to act

over the width of the trunnion assemblies.

Thus, if the ladle were supported at such points, the

ring stresses would become zero. Furthermore, the
straight stress lines in Figure 17 show that the ring
stresses are directly proportional with the distance "a,"
Figure 23, from the points discussed above to the actual
point of support on the trunnion pins.
A free body diagram, obtained by cutting along
section m-m, Figure 23, is shown in Figure. 24. If the
ladle were supported at a point on the vertical line
through the intersection of section m-m and the bottom
plate, the free body with zero ring stresses would be 'in
equilibrium. Now, while moving the point of support
the distance "a" to its actual location, ring stresses and
the shell shear "S" on the cut section are set up. Including the horizontal normal shell forces in the components
Ni of the ring forces at the cut section, equilibrium of
the free body requires the moment about the lower end
of the cut section to be zero:

C12= -g(a) Bik

C22= -g(a) [B 2k - P 2]
............. (27)
Ca2= -g(a) [Bak - P a]
C14= -f(a) Bik
C24= -f(a) [B 2k -'P 2]
............. (28)
Ca4= -f(a) [Bak - P a]
The moments and normal forces in the-stiffener rings
are then,
In the lip ring, at 0 degrees:

+ CI4)

rl (10c12
M I = 60

rl

F I= - 15 (10C12

(29a)

+ 4c14)

'

(30)

In the lip' ring, at 90 degrees:

rl (10c12 - CI4)
M I = - 60

(31a)

1
F I= - (lOc12 - 4c14)
15

(32)

In the top stiffener, at 90 dcgrees:

A similar free body diagram of the ladle cut along n-n,

Figure 23, furnishes another moment equilibrium condition:
2N Ih i

+ 2N 2h 2 + 2N ah a=0

(18)

Consistent with the assumed ring loadings it may be

shown that,

rl
.
(lOC22 - C24) +
60

M 2= - -

P 2r2 [1 - 0.6366 (a sin a

+ cos a) ]

F 2=15 (10c22 - 4C24) - P 2

60

Para [1 - 0.6366 (a sin a

4V2
6V2 C22 + 30
C24

. (20)

Similar expressions hold for rings 1 and 3.

The development of suitable formulas for the ring
loadings P 2, P a and the six c-coefficients now reduce to
a matter of solving Equations (16) to (20,). The resulting
expressions are given in Equations (26) to (28).
Summarized, the stress analysis procedure for circular
ladles includes the following steps. Use the auxiliary
expressions:
a=rt - 0.7071 r
e
(21)
rl

18

hI,

B2=E~2 h 2 , Ba=EIaa,h a
r2

ra

(22)

+ cos a) ]

(35a)

1
F.= 15 (10ca2 - 4Ca4) --; P s

(36)

The sign convention applied above is shown in Figure

21. Tensile normal forces and moments causing tension
at the outer ring surface are positive. The total ring
stresses are finally found by means of the well-known
formulas

II'.,,,.,..,.,..,

BI~EI~

(34)

In the lower stiffener, at 90 degrees:

M a= - ~ (10ca2 - Ca4)

N 2= -

(33a)

Ul

.A

Si

=---

Reprinted from

(37)

IRON AND STEEL ENGINEER, DECEMBER,

1949

where 0 stands for the outside and i for the inside ring
surface.
The procedure above applies for circular ladles. For
oval ladles, Equation (16) and consequently also
Equations (~6) to (28) * will include additional terms.
If the ladle cross section deviates little from circular,
as for most oval ladles, the effect of these additional
terms are small. The procedure will then coincide for
circular and oval ladles up to and including Equation
(!28).
The expressions for the moments in oval stiffener
rings subjected to the P-loads and skin shear loads are
slightly more complicated than those given above for
circular rings. Oval rings also sustain additional moments due to the internal liquid pressure. The oval ring
moments are then written, using the notation

l
u=rt
Oval lip ring, at 0 degrees:
M 1=

71'r
(10c12
60(7l'
2u)

a=rt - r sin 71'

2u

M 2=

~~ [10C (:~ :~)

22

+Pr
2

[1 _ 2 (ex
2

M 3=

V 2ur2
3
-ra
60

LIOc

32

3 -

(71'
71'

7l'

IRON AND STEEL ENGINEER, DECEMBER,

C24

(1l'; 2U) J......

-Ca4

(3Sb)

('ll")
]
71' + !2u

+ u) sin ex + ~ cos ex
71' + 2u

6 - 2U

.'(21b)

Round Ladle

--

+ u) sin ex + 2 cos ex]

71' + 2u

+ 4U)
+ 2u

(!2!Jb)

+ cP

C14

(1l') ]
1l' + !2u .(SIb)

6 - 2U 2)
71' + 2u

2 (ex

+ C14)

Figures 23 and 24 - The stiffener rings would have no

stresses if it were possible to apply the load to the
trunnions at a point inside the ladle. This is illustrated by moment arm "a" shown in the upper part
of the drawing (Figure 23). The lower part of the
drawing (Figure 24) Is a free body diagram taken
through section mom.

Reprinted from

rin~, a[t 90 de(~:s~u)

- IOc12 - - 60
71' + 2u

Oval top stiffener, at 90 degrees:

(38)

instead of Equation (!ill) gives better agreement with test results for
oval ladles. For u, see Equation (38). Oval lower stiffener, at 90 degrees:

Oval Ladle

Oval lip
M 1=

+ 2u

].

........ (35b)

In the formulas above, V is the resulting load per unit

length of the ring circumference. Roark's "Formulas for
Stress and Strain," Second 'Edition, p. 261, gives

+-yH 2 V2 . (39)
The expression for V 3 equals Equation (39) except forthe subscripts. VIis zero.
The normal ring forces F at 0 and 90 deg;ees a.re
identical for circular and oval rings neglecting the small
contribution from V.
In all the formulas given, the effect of the larger
Ipoments of inertia of the rings at the trunnions is
neglected. The expressions for moments and normal
forces in an oval ring with varying moment of inertia
are given in Reference (13). The effect of varying
moment of inertia is not large, and the exact formulas
would not appeal to a design engineer.
. From the interpretation of the experimental data, the
effective width of shell on each side of a ring, acting
together with the ring, was found to equal, at an
average, the width of the ring itself. Figure 25 thus
indicates the sections to be included in the cross section
areas and the moments of inertia of the rings. The
ring radius should be taken to the neutral axis of the
section. For the tests of ladle "e" less one or both
stiffeners (Table V), half of the thicker shell portion
was included in the moment of inertia where a stiffener
was removed.
The procedure outlined has been applied to the model
ladles, and the results are listed in Table V together
with the corresponding experimental stresses. The procedure fails in some cases to register the actual high
lip ring stresses at zero degrees and cannot be regarded
a5 perfect. To the best of the authors' knowledge,
1949

19

TABLE V
COMPARISON OF COMPUTED AND EXPERIMENTAL STIFFENER RING STRESSES
Full load - 2.5 in. hook distance
Locations
0: At trunnions
90: Between trunnions

Ladle "A"

Ladle "B"
8 x 16in.
Trunnions

8 x 8 in.
Trunnions

Outside surface
i: Inside surface

0:

-------------

Computed
psi

Measured
psi

Computed
psi

Measured
psi

- 7,200
+21,900

-10,800
+28,400

- 6,600
+20,000

- 6,900
+23,400

Computed
psi

Measured
psi

-------- -------

- 3,900
+ 7,900

Lip ring

+ 9,200
------

90

i
0

90

Top stiffener

i
0

90

Lower stiffener

+ 5,300
-14,100

+ 6,100
-20,900

+ 5,300
-14,000

+ 4,000
-18,000

+ 4,100
- 7,600

+ 2,800
- 6,300

+ 5,100
- 3,200

+ 4,400
- 2,800

+ 4,800
- 3,000

+ 3,500
- 2,000

+ 4,300
- 3,100

+ 3,400

+ 4,500
. - 2,400

+ 4,700
- 2,300

+ 4,000
- 2,600

+ 4,200
- 1,900

+ 5,400
- 5,100

+ 4,300
- 3,300

v
v

The discontinuity stresses in the side shell may be

found by means of Equations (1) to (9), and other
stresses checked by conventional methods.

Figure 26 - The diagram shows the dimensions of the

ladle used in the numerical example.

Other Ri"9S

Lip Ring

Figure 25 - The shell width which acts with the ring is

given in this diagram.

Xl.
o
o

::z:

o
~

however, it represents a distinct improvement over

methods now in use.
The ring stresses listed in Tables III and V correspond to about 2/5 of the stresses for a prototype ladle
loaded with molten metal.

NUMERICAL EXAMPLE

For a numerical stress analysis example, a round,

welded ladle of 150-ton capacity is chosen, Figure 26.
The ladle is meant to illustrate the procedure only, and
does not necessarily represent a desirable design. Ladle
height, top diameter, plate thicknesses and trunnion pin
diameter are chosen in accordance with the empirical
trend, Figure 18. The ladle capacity may be checked
by means of the formula
W r =1f"{ r 2H

3+ ~3 !(r
s
n

3r1i

+ 21'3)

(40)
I'lsi

where the radii should be taken to the inside of the

lining, which here is assumed to average 8 in. in thickness. With a maximum depth of 9 ft-6 in. of liquid
metal, and "{=420 lb per cu ft, Equation (39) gives
W=150 tons.
Only the stiffener ring stresses will be calculated here.

20

Reprinted from

IRON AND STEEL ENGINEER, DECEMBER, 1949

TABLE V - Continued
COMPARISON OF COMPUTED AND EXPERIMENTAL STIFFENER RING STRESSES
Full load - 7.5 in. hook distance

Full load - 2.5 in. hook distance

Locations
0 At trunnions
90: Between trunnions

With both
stiffeners

Outside surface
i: Inside surface

0:

---I

Computed
psi
0

Lip ring
0

90
Top
stiffener

90

Lower
stiffener

90

i
0

i
0

Measured
psi

t = 1.125 in.
h = 1.375 in.
e = 16.1 in.
r =62.06 in.
rt =71.75 in.

Less both
stiffeners

Less lower
stiffeners
Computed
psi

Measured
psi

Computed
psi

Measured
psi

Ladle "B"

Ladle "C"

--

8 x 16 in.
Trunnions
Computed
psi

Measured
psi

Both
stiffeners
Co'mputed
psi

Measured
psi

---

Computed
psi

Measured
psi

- 8,800 -18,500
- 5,300 - 9,200 - 8,700 -11,200 -18,900 -16,700 -11,000 -12,200 - 7,200
+11,100 +Vield +17,100 +Vield +37,400 +Vield _+33,300 +33,600 +14,800 +17,900 +18,500 +Vield
+ 3,900 + 3,500 + 6,300 + 4,700 +13,800 +10,100 + 8,800 + 7,700 + 7,600 + 5,700 + 6,500 + 6,000
- 8,100, - 5,900 -11,400 - 7,700 -24,900 -17,200 -23,900 -23,800 -14,100 - 7,500 -13,500 - 9,500
+ 8,200 + 5,300 + 8,400 + 6,700 + 5,900 + 1,700 + 8,000 + 5,600 + 8,000 + 6,400 +13,600 + 9,300
- 5,200 - 3,400 - 5,000 - 4,600 - 6,200 - 2,700 - 5,000 - 3,100 - 5,700 - 4,20~_ - 8,600 - 5,700
+ 4,800 + 4,200 + 1,600 + 2,000 + 3,500 + 3,600 + 6,500 + 7,600 +10,100 + 8,400 + 8,000 + 4,400
- 3,000 - 2,300 - 1,600 - 1,300 - 3,700 - 2,700 - 4,300 - 4,200 ~ 9,500 - 7,000 - 5,000 - 4,400

The dimensions given in Figure

II =10.4 in. 4
12 =86.5 in. 4
S1O= 8.4 in. 3
S20 = 24.6 in. a
Sli = 3.8 in. 3
S2; =38.9 in. a
2
A 1=20.4 in.
A 2=28.0 in. 2
hi =136 in.
rl = 73.:52 in.

-------

Ladle "A"

Ladle "C"

h 2 =105 in.
r2 71.32 in.

26 yield
1 3 =372.0
Sao = 63.3
S3i = 87.7
A a= 37.1

in. 4
in. a
in. a
in. 2

h a = 5,15 in.
ra =69.17 in.

E = 30,000,000 psi
w =56.0 in.
sin a=~=0.1300
6rt
cos a = 0.991:5
a =0.1304-

Equation (34): F 2 =' -, 14,OOOlb

Lower stiffener, no degrees:
Equation (35a): M a=
1,033,000 lb in.
Equation (36): Fa =
- 35,000 Ib
By disregarding the lip ring
C12 = C22 = C32 = C14 = e24 = ea4= 0
and
Figure 27 - The graph shows numerical values of g(a)
and f(a) as functions of the ratio of trunnion width w
to the side shell radius, rt atthe trunnion level.
9 (a) =4.5 [0.5708-a+ (2-sina-cosa)sina]
f(a) =225 [2.5708-a+ (2+sina-cosa)sina
-2.5464(asina+cosa)]
where a = sin

Capacity, Equation (39), We

=300,000 lb
Steel dead'weight
W. = 0.19We = -.57,000 lb
WL=0.85 W. = 48,0001b
Lining
W -=
405,000 lb
Equation (21): a
= 43. 13 in.
Equation (22): B 1 =106,000Ib
Ih=751,000 lb
13 a= 1,855,000Ib
Equation (23): k
= 0.0~24
Equation (24): g(a) = 2.496
Equation (25): f(a) = 4.590
The ring loadings are:
Equations (26): P 2= 18,100 Ib
P a=44,800 lb
Equations (27):
C12 = - 5,900 lb
e22 = 3,300 lb
ea2 = 8,200 Ib
, Equations (28):
e14= - 1O,8001b
e24=6,200 Ib
ca4~'1.5,100 Ib
The moments and normal forces in the rings due to
these loadings become:
Lip ring, 0 degrees:
Equation (29a): M 1=
85,000 lb in.
Equation (30): F 1 =
7,0001b
Lip ring, 90 degrees:
Equation (3b) M 1=
59,000 lb in.
Equation (32): F 1 =
1,0001b
Top stiffener, 90 degrees: Equation (33a): M 2=
430,000 Ib in.
Reprinted from

IRON AND STEEL ENGINEER, DECEMBER, 1949

21

TABLE VI - CALCULATED CRITICAL RING STRESSES

IN LADLE, FIGURE 26

Ring

Surface
0

Lip ring ...... , ' , ,

i
0

Top stiffener, , , . , '

i
0

Lower stiffener, . , .

Maximum stress Maximum stress

with lip ring,
without lip ring,
psi
psi
- 9,800
" +22,700
+ 17,000
-11,600
+ 15,400
-12,700

........
........

+ 18,200
-12,500
+ 16,300
-13,800

P 2 = 18,100 lb
P a=44,800 lb
as before the corresponding moments and normal forces
in the stiffeners are, Top stiffener, 90 degrees:
Equation (33a): M 2 =
463,0001b in.
Equation (34): F 2 =
-18,000Ib
Lower stiffener, 90 degrees
Equation (35a): M a=
1,109,000 lb in.
Equation (36): Fa =
-45,000Ib
The unit stresses in the' two' cases, i.e. including and
disregarding the lip ring, are now calculated by means
of Equations (37). Table VI gives the resulting maximum stresses in each ring. The lip ring usually is considerably weaker than the other stiffeners and will show
\ltresses which are nearly proportional to the depth of
the ring. To avoid high stresses, as obtained in this.
example, the lip ring "hould be made even narrower
than shown in Figure 26. Experience shvw", on the
other hand, that a yield stress in the lip ring does not
endanger the safety' of the ladle.

SUMMARY

Structural tests have been made of three 1/5 scale

models of 150-net ton capacity hot-metal ladle prototypes. The results of the experimental and theoretical
study are, in brief:
.
1. The structural behavior of hot-metal ladles deviates
considerably from what has often earlier been
assumed.
2 Under load, the ladle side deflections vary approximately linearly from zero at the bottom to the
maximum at the lip. This deflection is inward in
the trunnion region and outward between the
trunnions.
3 The stiffener rings obtain the highest stresses at the
trunnion line and midway between the trunnions.
The lip ring carries the largest stresses, and the
lower stiffener the smallest.
4 The side shell suffers onl;)' small stresses, except
near the bottom juncture, where large discontinuity
stresses are found.
.5 Flat bottoms have very large deflections, and the
stresses both at center and edge are likely to exceed
the yield point.
6 Dished bottoms show moderate deflections and
stresses"and are superior to the flat types.
7 Stresses and deflections increase nearly ,linearly
with both amount of load and distance between the
supporting ladle crane hooks, except for bottom

22

stresses which are not affected by the hook distance.

8. Tilting of the ladle does not increase the stresses.
9. No major difference in structural behavior was
found between riveted and wclded ladles.
10. No major difference in structural behavior was
found between circular and oval ladles.
11. Wider trunnion assemblies tend to reduce the
stiffener ~ing stresses.
12. A spacer band or a heavy middle shell portion
reduces the stiffener ring stresses approximately in
proportion to the dead weight added by these
members.
13. Although the load-carrying contribution of the
stiffener rings increase with their distance from the
bottom, the lip ring should not be -emphasized in
ladle design due to its exposure to damage. Design
stresses should be calculated both considering and
disregarding the lip ring.
14. Stress analysis procedures which give a fair agreement with the experimental res,ults are proposed
for stiffener rings, flat bottom plates, and the
bottom-side juncture discontinuity stresses. A
numerical design example is given. When finished.
the work of the design division of the Pressure
Vessel Research Committee may furnish a design
procedure for dished bottoms.
15. Summaries of design practices and recommendations
as given in various articles on hot-metal ladles are
incorporated in this report.

NOMENCLATURE

Subscripts)
_
i= 1, 2, 3 for rmgs and shell fields as shown
Figure 21.
0= outside ring surface.
i = inside ring surface.

III

Roman alphabet)
a
Moment arm of trunnion reaction: (Figures 23
and 24.)
Ai
Area of effective stiffener ring cross sections.
Ai' Net areas of stiffener cing cross sections, including
only that shell in contact with ring.
Bi
Stiffener ring constants given by Equation (22).
Cij
Coefficients of shell shear flow. Equation (14).
First subscript refers to ring number, second to
sine series term number..
d
Depth of dished head.
e
Distance along trunnion center line from center
of supporting ladle crane hook to middle of side
shell. (Figure 23.)
E
Youngs modulus.
f(a) A function of a given by Equation (25).
Fi
Normal force in rings.
g(a) A function of a given by Equation (24).
h
Bottom plate thickness, cover plate excluded.
h,
Vertical distance from rings to bottom juncture.
H
Total maximum depth of liquid metal.
Hi
Head of liquid metal above rings.
Moment of inertia of rings.
Ii
k
An auxiliary constant given by Equation (23).
I
Half the length of straight shell section of oval
ladle. (Figure 21.)
Reprinted fl"Om

IRON AND STEEL ENGINEER, DECEMBER, 1949

Moment in rings.
Moment per inch along juncture of side shell and
bottom.
M x Moment per inch of vertical section through side
shell, or circumferential section through bottom
plate.
My Moment per inch of horizontal section through
side shell or radial section through bottom plate.
Components of normal ring force at = 45 degrees.
Equation (19), (Figure 24).
Ni' Components of normal ring force at = 45 degrees.
Equation (20).
Liquid pressure per unit area of bottom plate.
Transverse shear on side shell per inch of juncture
between side shell and bottom.
Pi
Concentrated ring loads. (Figure 21.)
qi
Shear flow per unit length of horizontal sections
through side shell.
r
Bottom plate radius.
rj
Radii of neutral axes of rings.
rH
"Wet" radius at liquid metal surface.
rt
Shell'radius at trunnion level.
R
Radius of curvature of hemispherically dished
bottom.
S
Slope of ladle side.
Si
Section modulus of rings.
t
Side shell thickness.
u
Constant for oval ladles. Equation (38).
Width of stiffener rings. (Figure 25.)
Liquid pressure per inch of circumference of rings.
Equation (39).
w
Width of trunnion assembly. (Figure 22.)
W
Total weight of loaded ladle.
We Ladle capacity. Equation (40).
y
Distance along side shell meridian from bottom
juncture.
Greek alphabet
a
Half angle between concentrated rmg loads.
(Figure 22.)
(3
Side shell constant. Equation (1).
Specific weight of liquid metal.
'Y
v
Poisson's ratio, taken as 0.3.
In ring analysis, an angle, Figure 21.
=3.1416.
Normal stress per unit area.
In ring analysis, an angle, Figure 21.
1

DISCUSSION
PRESENTED BY

L. LARSON, Chief Engineer, Corrigan-McKinney

Gives list of manufacturers, number and types of welded ladle~

madc since 1932 and where located. Discusses savings in weight and
some causes of cracks in the plates.
3. "Ladle Design and Service," by J. H. HRUSKA, "Blast Furnace
and Steel Plant" v. 19, 1931, pp. 673-677 and 836-838.
Discusses the design and maintenance of steelworks ladles. Information is given concerning types of slag taken from ladles used for
different classes of steel;'the properties of ladle linings; the principal
dimcnsions of refractory parts; the chemical composition of stopper
parts and ladle nozzles; discusses methods of drying linings; emphasizes the importance of supporting parts.
4. "Welded Ladles for Open-Hearth Service," by F. L. LINDEMUTH,
"Steel" v. 109, 1941. pp. 74-77.
.
Prcsents data on the design of round and elliptical welded steel
ladles with capacities of 50-175 tons of moltcn steel. It is pointed out
that with welded construction the weight of the ladle is about onesixth that of the contents, whereas for riveted ladles the weight ratio
is 1:4.
5. "Welded Ladles Help Increase Steel Output," by G. R. REISS,
"Steel" v. 112, 194~. pp. 112, 115.
By fabricating ladles from welded steel plates weight was reduced
by over one-third, obtaining greater capacity without increasing total
load on cranes.
6. "Standard Design Specifications for Welded Steel Ladles,"
Carnegie-Illinois Steel Corp., Operating Dept., Engineering Division.
S. P. No.2, Jan. 10, 1945. Unpublished.
7. "Lining and Sehedrding Ladles for Maximum Life," by L. R.
BERNER, A.I.M.M.E., Proc. of Open Hearth Comm., 1945, p. 153.
8. "Hot Metal Ladles of Spherical Shape," by F. L. KLING, "Iron.
Age" Aug. 1, 1929, p. 284.
Description of the "Kling Ladle."
9. "Blast Furnace Practise (1929)," by F. CLEMENTS. Vol. 2:
Design of Plant and Equipment, pp. 401-403.
Illustrated descriptions of hot-metal ladles and cars.
10. "Ermittlungen Des Kippmamentes Und Der Kippkraft Einer
Giesspfanne," by B. OSANN, "Giesserei-Zeitung" v. 2'2, 1925, pp. 44-46.
Determination of the tipping moment and tipping force of a ladle.
Calculation of center of gravity; behavior of ladle at interruption of
peration; tipping force at different tipping positions.
11. "The Basic Elastic Theory of Vessel Heads Under Internal
Pressure," by G. W. WATTS and W. R. BURROWS, Standard Oil Co.,
Indiana, March 1949, p. 55, ASME Journal of Applied Mechanics,
vol. 16, No. 1.
Means are outlined for calculating stresses and deformations at or
near the junctures of vessel shells with flat, ellipsoidal and other type
heads.
1'2. "The State of Stress in -Full Heads of Pressure Vessels," by
W. M. COATES, ASME Trans. v. 52, 1930, APM-52-12.
13. "Structural Analysis of Hot-Metal Ladles," by K. E. KNUDSEN,
October 1948. Dissertation. Unpuhlished. Available for loan at Fritz
Laboratory, Lehigh University.
Presents the development of the formulas given in this report and
discusses other methods of attacking the stress analysis pr::>blem.
14. "Formulas for Stress and Strain (1948), by R ..J. ROARK,
McGraw-Hill Book Co.

In shell analysis, functions are given b.y Equation

(6).

BIBLIOGRAPHY
1. "Structural Tests of Hat-lyJetal Ladles," by K. E. KNUDSEN,
W. H. MUNSE and B. G. JOHNSTON. Report to AISE, Standardization
Committee, January 1948. Unpublished. Available for loan at Fritz
Laboratory Lehigh University.
The results of a study of the structural behavior of hot-metal ladles
are offered. Three 1/5 scale stcel model ladles were tcsted for strains
and deflections under static loads, disregarding the temperature
effects of the molten metal. The models were: a round riveted ladle,
an oval welded ladle, and a round welded ladle, all bascd on a prototype of 150-ton net capacity. The test results are given in the form
of deflection diagrams, stress diagrams, and tables. The principal
stresses as well as horizontal and vertical strcss components, as
prescnted, have bcen evaluated from the measured strains.
2. "History and Use of JVelded Ladles," by F. L. LINDEMUTH,
A.I.M.M.E. Iron and Steel Division, Open Hearth Comm., Open
Hearth Proc. 1938, pp. 27-32.

Reprinted from

IRON AND STEEL ENGINEER, DECEMBER, 1949

Plant; Republic Steel Corp., Cleveland, Ohio

F. LINDEMUTH, Chief Engineer, William ~. Pollock Co., Youngstown, Ohio
B. G. JOHNSTON, Director, Fritz Engineering
Laboratory, Lehigh University, Bethlehem, Pa.
K. E. KNUDSEN, Fritz Engineering Laboratory,
Lehigh University, Bethlehem, Pa.
W. H. MURSCH, Staff Engineer, Carnegie-Illinois
Steel Corp., Pittsburgh; Pa.
W. W. TRINKS, Associated Engineers, Pittsburgh, Pa.
J. A. EVANS, Assistant Chief Draftsman, Republic
Steel Corp., Canton, Ohio

Leonard Larson: I believe that this program, as'

outlined and summarized in Mr. Knudsen's paper, is a
fundamental step toward a clearer concept of what takes
place in a ladle under load. For many of us, there have

23

been too many fluid spots in attempting to' analyze

stresses in ladles. In many cases we have probably
treated these blind spots with our best judgment, which
at times may not have been very sound. The fact that
there have been so few mishaps with hot metal ladles
indicates that at least we have exercised our ignorance
on the safe side.
Tests to date have been made without hot metal in
the ladle, so we stUI have one major blind spot, namely
influence of heat from hot metal on ladle behavior.
F. Lindemuth: The ladle models had complete li~
rings for retaining the lining at the top of the shells. In
the tests these lip rings supplied a large percentage of
the strength of the models. Mr. Knudsen has said in
his discussion that in actual practice this lip ring should
not be included in any strength calculations, and this
point should be emphasized.
Member: I make a few ladles and I am quite interested in whether there are any records of failures. In
other words, with all of these figures that we have, can
we apply those to any ladles that have failed, except
for heating and overloading? 1 know they are overloading ladles. We design a ladle for 180 tons and the first
thing you know they have added to the top and it is a
. 225-ton ladle.
Bruce Johnston: We think it is of considerable
intere~t, that all of the design procedures we were able
to lay our hands on do not even give the correct sign
of the stress. That is, where there actually is compression
stress, the design procedure may indicate tension stress.
The method that Mr. Knudsen has developed in
every case gives the correct sign of the stress and in
most cases is within 10 or '20 per cent of the magnitude,
so with these formulas you can be confident in the
design of the ladle and know that the rings, at least,
.
will not be overstressed.
Member: A question that comes into our operations
a good many times is what arc the steel companies to
accept in the way of a factor of safety in ladles?
I think the answer to that, as far as we have figured
out, is that we used to build the riveted ladles with a
factor of safety of ten to fifteen and then, when cranes
were being .overloaded the users wanted lighter ladles
so the steel companies' began to drop their factor of
safety.
My question is, how far can you carry that? It seems
to me this committee should make some recommendations on what the factor of safety should be in a ladle.
Leonard Larson: I think the answer to that is
arbitrary. That is, there would be as many opinions as
there are individuals discussing it. However, certainly
their ability to use better judgment has been sharpened
by more information.
K. E. Knudsen: The last discussion mentioned the
change from riveted to w~lded ladles. I should have
mentioned, of course, we just have one riveted ladle
and just two welded ones, so there is no broad basis for
conclusions, but we did not find any basic difference in
structural behavior.
Member: I think the answer to that is that you were
working on the basis of a modern riveted ladle and a
.modern welded one. However, you go back ten or fifteen
years and examine some of the riveted ladles we built
ten or twenty years ago and you will find the factor of
safety is much higher all of the way through.

24

K. E. Knudsen: That question, too, could not be

decided from these tests at all. You have to take into
consideration the temperature effect, of which we know
very little.
Bruce Johnston: In connection with the factor of
safety, I might say that you cannot know what your
factor of safety is unless you know what your stresses
are. Therefore, this report should provide a step toward
a real factor of safety instead of a "factor of ignorance."
Leonard Larson: In considering factor of safety as
related to objective of reducing. excess weight in hot
metal ladles by more intelligent design, it seems obvious
to me that we must continue to provide for some abuse,
and normal deterioration due to wear and tear. Where
inspection and maintenance are below par, a higher
factor of safety is necessary as compared with a higher
level of inspection and maintenance. It is difficult to
draw a definite line between economy and safety, where
so many variables may be involved.
Wm. Mursch: In our office there are a good many
specifications coming through for ladles. You look 'at
these different specifications and drawings and see the
different types of ribs and thicknesses of plates specified
on these drawings for the same size ladles and you
wonder which is the correct design.
In other words, for a 150-ton ladle one plant will have
a bottom plate specified maybe an inch and an eighth
thick and the same size ladle from another plant will
specify an inch and a half thickness. The result is that
we have never had any real criterion to tell what is a
good design or a poor design. The question is always
raised, which is right and which is wrong. Now, these
formulas which have been presented in this paper will
help in checking and in getting a ladle which is designed
more theoretically correct than what we have had in
the past.
W. W. Trinks: I would like to ask whether, in the
test ladles, there was a lining?
K. E. Knudsen: Yes, we had one. on all tests with a
thickness of about %: in. on the side, and one in. on
the bottom.
W. W. Trinks: This whole investigation was originally instituted, for the purpose of finding out whether
welded ladles could be made lighter than riveted ladles.
Did you reach a conclusion on that?
K. E. Knudsen: It was not the specific purpose; it
was one of the purposes of the investigation. Although
we had riveted and welded ladles, we did not find any
major difference in stress behavior between the two.
J. A. Evans: Did you check the stresses by any
known methods, such as strain gages or photo-elastic
devices?
K. E. Knudsen: We used bonded electric strain gages.
Bruce Johnston: How many strain gage readings
did you make?
K. E. Knudsen: On each ladle we had approximli,tely
150 of these gages and we took in all about 14,000
read~ngs measured by means of these electric strain
gages.
Bruce Johnston: This last table he showed: was
the check between the strain readings and the method
of analysis. Mr. Knudsen tried over thirty different
empirical assumptions in arriving at the best check of
all thrfle ladles.

Reprinted from

IRON AND STEEL ENGINEER, DECEMBER,

1949